How Do Dyes Impart Color to Different Components of the Tissues?

Chapter 19 | On Chemical Reactions and Staining Mechanisms

John A. Kiernan, MB, ChB, PhD, DSc

A Tailpiece: Hematoxylin & Eosin (H & E)

Further Reading

A special case, though not a special stain. H & E is so widely used

ettinger C & Zimmermann HW (1991) New investigations on hematoxylin,

B
hematein and hematein-aluminium complexes. 2. Hematein-aluminium
complexes and hemalum staining. Histochemistry 96: 215-228.

that no account of staining mechanisms can fail to mention it the

more so as H & E still puzzles us in various ways.
What is H & E? The easy bit is to say what eosin is. Eosin
Y is a normal acid (anionic) dye of moderate size and well studied
chemistry. Hematoxylin however is a misnomer, hematoxylin being
a colorless natural product whose oxidation can produce hematein.
And hematoxylin in H & E is the product of the reaction of hematein

Horobin RW (1982) Histochemistry: an explanatory outline of histochemistry

and biophysical staining. Fischer: Stuttgart & Butterworth: London.
Horobin RW & Bancroft JD (1998) Trouble-shooting histology stains.
Churchill Livingstone: New York.
Horobin RW & Bennion PJ (1973) The interrelation of the size and substantivity
of dyes: the role of van der Waals attractions and hydrophobic bonding in
biological staining. Histochemie 33, 191-204.

What is the difference between argentaffin

and argyrophilic reactions?
Both terms relate to the formation of colloidal metallic silver (Latin
argentum, Greek argyros) (1), which may range in color from yellow
through brown to black depending on particle size and density, in
specific structural components of tissues. The metal is formed at
sites of reduction of silver ions derived from a staining solution. These

The only argentaffin reaction of importance in histopathology is the

reduction of an alkaline solution containing silver diammine in the
Masson-Fontana stain, which detects small quantities of melanin
and, in suitably fixed tissues, the amines derived from tyrosine and
tryptophan: dopamine, noradrenaline, adrenaline and serotonin (6).
Most types of melanoma and some tumors of intestinal endocrine
cells give positive Masson-Fontana argentaffin reactions (7, 8).

ions may be simple Ag+, as in a solution of silver nitrate, or they

Nearly all the cell-types of the diffuse endocrine system (DES) are

may be complex ions such as silver diammine, [Ag(NH3)2]+ or silver

argyrophilic (9, 10), as are the axons of neurons and reticulin, a term

methenamine, [Ag3(C6H12N4)2]3+, which are formed respectively by

that applies to thin collagen fibers and basement membranes. The

adding ammonium hydroxide or hexamethylenetetramine to solutions

technical details of argyrophil silver methods vary with the structures

of silver nitrate. (Hexamethylenetetramine, the reaction product of

for experimental data and summary of prior work see Bettinger

to be demonstrated, but in all cases there are at least two steps. In

formaldehyde and ammonia, is also used as a urinary antiseptic;

& Zimmermann (1991) and Puchtler et al (1986). Moreover the

the first, which is often called impregnation, the tissue (whole pieces

its pharmaceutical name is methenamine in the USA and hexamine

Al-hematein complex present in tissue sections following blueing

or sections) is exposed to a silver-containing solution. Silver ions bind

in the UK.) These complex silver ions are more easily reduced than

is probably different again, perhaps a poorly soluble olation polymer

non-specifically to the tissue and at certain significant sites some

Ag+. By convention, argentaffin is used when the reducing agent is

(Puchtler et al 1986).

of the ions are reduced to silver nuclei, each consisting of several

present in the tissue. Structures are said to be argyrophilic when

atoms of the metal and far too small to be visible. In the second step,

bright light or a solution of a reducing agent is applied to the object

known as development, a reducing agent changes dissolved silver

after exposure to a silver solution.

ions (which may or may not be derived from those non-specifically

with aluminium ions, so is better termed Al-hematein. But to ask

what is this? is to encounter puzzle number one. The nature of the
Al-hematein complex or complexes in staining solutions remains
uncertain. Both cationic and anionic complexes could be present;

How does H & E work? Again starting with the easy bit: eosin
stains sections by acid dyeing. The Al-hematein staining mechanism

Lyon H (1991) Theory and strategy in histochemistry. Springer-Verlag, Berlin.

Prent P (2009) Staining of macromolecules: possible mechanisms and
examples. Biotech Histochem 84: 139-158.
Puchtler H, Meloan SN & Waldrop FS (1986) Application of current chemical
concepts to metal-hematein and brazilein stains. Histochemistry 85: 353-364.

however is puzzle number two. Its staining behavior differs from

An example of an argentaffin reaction with silver nitrate as the reagent

nuclear and Nissl staining basic dyes such as cresyl violet or toluidine

is the histochemical demonstration of ascorbic acid in such sites as

blue. Extraction of DNA from cell nuclei inhibits nuclear staining by

the adrenal cortex and the embryonic eye. Tiny fragments of fresh

such basic dyes, but does not eliminate Al-hematein staining. Routine

tissue are immersed in an acidified solution of silver nitrate and

basic dyes readily wash out of sections if over-enthusiastically

then in a solution of sodium thiosulfate (2, 3, 4). The latter reagent

dehydrated by alcohol, but hematoxylin does not. Nuclear staining

removes unreduced silver, which otherwise would be slowly reduced

by hemalum after extraction of DNA may be attributable to histones,

by the action of light, discoloring all parts of the tissue. This slower

the strongly basic proteins of eukaryotic nuclei. It has been argued

reduction, which forms the basis of several traditional methods for

(but not proved) that hemalum solutions contain anionic dye-metal

showing intercellular clefts and spaces (5), is an argyrophil, not an

complexes that would be attracted to the protonated lysine and

argentaffin reaction, because the light serves as an external reducer,

arginine side-chains of histones. It has also been suggested that

acting probably on silver chloride precipitated in extracellular fluid.

Al-hematein is a large hydrophilic basic dye, with resistance to alcohol

The same holds for von Kossas stain for calcified material. In this,

extraction due to insolubility in that solvent; an alternative speculation

silver ions from aqueous AgNO3 displace calcium from deposits of

notes that olation polymers are typically insoluble.

Conclusions H remains a puzzle in some respects, although
E is easy. Turning speculations into knowledge will require renewed

bound by the tissue) into deposits of colloidal metal. This development

reaction is catalyzed by the silver nuclei, which are thereby enlarged
enough to form visible deposits. Thus, the second (development) step
of an argyrophil method is a chemical amplification of the invisible
signal generated in the first (impregnation) step (11). Two examples
of argyrophil methods will serve to illustrate different determinants of
specificity and types of catalytic amplification.

calcium carbonate or phosphate, forming insoluble silver carbonate

or phosphate, which is quickly decomposed by the action of light to
form black colloidal silver.

chemical investigation of the actual Al-hematein complexes present

in the dyebaths, and of the actual nature of blued hematoxylin.

166 | special stains and h & E

special stains and H & E

| 167

On Chemical Reactions and Staining Mechanisms

On Chemical Reactions and Staining Mechanisms

In the Grimelius technique, for showing cells of the diffuse endocrine

References

system (10), sections are impregnated with 0.004 M AgNO3 in

1. Shorter Oxford English Dictionary, 3rd ed. Oxford: Clarendon Press,

1973, pp. 102, 103.

a 0.02 M acetate buffer at pH 5.6 for three hours at 60 C. This

solution has been found empirically to provide for nucleation of
silver in and around the secretory granules of cells of the DES.
The slides are then transferred to a 0.09 M solution of hydroquinone
in 0.4 M sodium sulfite, in which they remain for one minute. Sulfite
ions react with silver ions (partly derived from residual impregnating
solution and partly from silver non-specifically bound to the tissue)
to form a complex anion, [Ag(SO3)2] 3. This is reduced by hydroquinone and the resulting black colloidal silver accumulates around
the initially formed nuclei in the cytoplasms of cells of the DES. A
negative result with this simple, inexpensive technique may avoid
multiple expensive immunohistochemical tests for the various
peptides and proteins that characterize different DES cell-types and
tumors derived from them (8).
There are several silver methods for reticulin that have in common
three major steps (12). Gomoris technique (13) is typical. The
sections are treated for one minute with 0.06 M aqueous potassium
permanganate. This oxidizes the neutral sugar components of
glycoproteins to aldehyde groups. (Reticular fibers and basement
membranes contain relatively more carbohydrate than larger collagen
fibers.) Brown manganese dioxide is deposited all over the sections
and is removed by rinsing in 0.09 M potassium metabisulfite. After
washing well with distilled water the sections are placed for two
minutes in an alkaline silver diammine solution (approximately 0.4 M
total Ag). This reagent is reduced to metal by aldehydes and some

2. Bourne G. Vitamin C in the adrenal gland. Nature 1933;131:874.

3. Willis RJ, Kratzing CC. The chemistry of the silver precipitation method
used for the histochemical localization of ascorbic acid. Stain Technol
1974;49(6):381-386.
4. Lam K-W, Zwaan J, Garcia A, Shields C. Experimental detection of
ascorbic acid in the eye of the early chicken embryo by silver staining.
Exp Eye Res 1993;56:601-604.
5. Lee AB. The Microtomists Vade-Mecum. A Handbook of the Methods
of Microscopic Anatomy. London: Churchill, 1885, pp. 101-104.

How does pH influence the staining

mechanism of Schiff's reagent in biological
tissue samples?
Schiffs reagent is a solution made by reaction of basic fuchsine, a red
dye, with sulfurous acid. The reagent ideally is colorless, but it may be
light yellow on account of impurities in the dye. It was introduced by
Hugo Schiff in 1866 as a chromogenic test for aldehydes. Traditionally,
basic fuchsine was a mixture of four cationic red triphenylmethane
dyes (pararosaniline, rosaniline, magenta II and new fuchsine) made

whereas products sold as basic fuchsine typically are mixtures

(2, 3). All four components of basic fuchsine are suitable for making
Schiffs reagent, but pararosaniline has been the one used in chemical
studies. Triphenylmethane dyes are commonly formulated as shown
in Figure 1 (I - IV) with the positive charge on a nitrogen attached to
a quinonoid ring, but other resonance forms are equally valid. The
carbonium ion structure (V) allows for the simplest explanations of
the formation and reactions of Schiffs reagent. The balancing anions,
not shown in Figure 1, may be chloride or acetate.

6. Lundqvist M, Arnberg H, Candell J, et al. Silver stains for identification of neuroendocrine cells. A study of the chemical background.
Histochem J 1990;38:615-623.

by oxidizing a crude mixture of aniline and mixed ortho- and para-

The reaction with sulfurous acid (Fig. 1, V) decolorizes the dye. The

isomers of toluidine (1). Pure aniline, p-toluidine and o-toluidine

structure of the colorless product was debated for many years (4),

7. Grizzle WE. Theory and practice of silver staining in histopathology.

J Histotechnol 1996;19(3):183-195.

have been articles of commerce for many years. Consequently,

but it is now generally agreed to be an alkylsulfonic acid as shown

pararosaniline and new fuchsine are available as individual dyes,

(for Schiffs reagent made from pararosaniline) in Figure 1 (5, 6, 7).

8. Grizzle WE. Silver staining methods to identify cells of the dispersed

neuroendocrine system. J Histotechnol 1996;19(3):225-234.
9. Montuenga LM, Guembe L, Burrell MA, et al. The diffuse endocrine
system: from embryogenesis to carcinogenesis. Prog Histochem
Cytochem 2003;38(2):155-272.
10. Grimelius L. Silver stains demonstrating neuroendocrine cells. Biotech
Histochem 2004;79(1):37-44.
11. Gallyas F. Physicochemical mechanisms of histological silver staining
and their utilization for rendering individual silver methods selective
and reliable. Biotech Histochem 2008;83(5):221-238.
12. Lillie RD, Fullmer HM. Histopathologic Technic and Practical Histochemistry. 4th ed. New York: McGraw-Hill, 1976, pp. 683-692.
13. Garvey W. Some favorite silver stains. J Histotechnol 1996;19(3):
269-278.
14. Velican C, Velican D. Silver impregnation techniques for the histochemical analysis of basement membranes and reticular fiber networks. In:
Glick D, Rosenblaum RM, eds. Techniques of Biochemical and Biophysical Morphology. v. 1. New York: Wiley, 1972: pp. 143-190.

of the complex silver ions are nonspecifically bound to the tissue.

After a brief rinse in water the slides are placed in 20% formalin
(2.7 M formaldehyde) for 4 minutes. This reacts with residual and
loosely bound silver diammine ions and the resulting silver is
deposited preferentially upon the catalytic silver nuclei formed at
the sites of aldehyde groups produced by the initial permanganate
oxidation (14).

Figure 1. The components of basic fuchsine (I - IV) and the formation of Schiffs reagent by reaction of pararosaniline with sulfurous acid (V,VI).

168 | special stains and h & E

special stains and H & E

| 169

On Chemical Reactions and Staining Mechanisms

Studies with NMR spectroscopy indicate that the major component

On Chemical Reactions and Staining Mechanisms

be stable while exposed to the air and losing sulfur dioxide for 15

now contains some regenerated basic fuchsine, which could bind to

changes when slides are washed after immersion in Schiffs reagent.

is the zwitterionic form shown (VI) in which one of the amino groups

to 20 minutes. Accordingly, Schiffs reagent for use as a biological

polyanions in the tissue and give false-positive staining in cartilage

Rinses in 0.03 M sulphurous acid to remove unbound reagent were

is protonated (6).

stain must be acidic (pH about 2 is best for most applications) and it

matrix, mast cell granules and other materials that are stainable by

formerly commonplace (10, 11) but they weaken the final color. Rapid

must contain more sulfurous acid than was needed to decolorize the

cationic dyes at low pH. Reaction VI V of Figure 2 is reversible,

washing in water does not result in false-positive staining (12) and

basic fuchsine. If the pH approaches 3 a white precipitate may form

so Schiffs reagent that has become pink from losing SO2 can be

is currently recommended (9, 13). The rise in pH associated with

of an excess of sulfurous acid. Raising the pH or reducing the

after a few months of storage (4, 9). In equation VII it is seen that a

regenerated by adding a source of sulfurous acid, such as thionyl

washing in water for 10 minutes reduces the protonation of the stable

concentration of sulfurous acid increases the sensitivity of the

solution at pH 2 contains predominantly bisulphite ions and sulfurous

chloride or sodium metabisulphite. Excessive acidification causes

dye-tissue complex (X in Figure 2), with resultant intensification of

reagent for detecting aldehydes but also leads to regeneration of the

acid molecules. Reversal of the sulfonation, VI V, occurs if there

additional protonation, leading to the formation of IX, which does not

the color. The same effect can be achieved in 15 seconds with 0.3%

original dye. Some of the equilibria that account for this instability

is loss of SO2 into the atmosphere (see VII) or if the pH is raised

react with aldehydes (6, 8).

sodium tetraborate, an alkaline solution (14).

(6, 8) are summarized in Figure 2 (VII, V-VI). For histochemical

towards neutrality. SO2 can escape from a loosely capped bottle. The

purposes high sensitivity is not usually needed and the reagent must

Schiffs reagent then becomes pink and is unfit for use because it

The colorless solution is stable only at low pH and in the presence

The simplest (theoretically) and oldest way to make Schiffs reagent

is to bubble gaseous sulfur dioxide through a 0.3-1.0% (0.01-0.03M)
aqueous solution of basic fuchsine. This is the method implied in
Figure 2. The acidity of the solution is then due only to ionization of
sulfurous acid (pK1 = 1.9) but it is almost impossible to adjust the
amount of added SOCI2 in the range between the minimum needed
to decolorize the dye and the maximum dictated by the solubility of
SOCI2 gas in water.Another technique is to add some thionyl chloride,

1. Cooksey C, Dronsfield A. Fuchsine or magenta: the second most

famous aniline dye. A short memoir on the 150th anniversary of the
first commercial production of this well known dye. Biotech Histochem
2009;84(4):179-83.
2. Horobin RW, Kiernan JA, eds. Conns Biological Stains. A Handbook
of Dyes, Stains and Fluorochromes for Use in Biology and Medicine.
10th ed. Oxford: BIOS, 2002, pp.184-186, 190-191.

SOCl2, to the dye solution. This liquid reacts instantly with water:

3. Green FJ. The Sigma-Aldrich Handbook of Stains, Dyes and Indicators.

Milwaukee, Wisconsin: Aldrich Chemical Company, 1990, pp. 126-130.

SOCl2 + 2H2O H2SO3 + 2H+ + 2Cl providing both sulfurous

4. Kasten, FH. The Chemistry of Schiffs reagent. Int Rev Cytol 1960;10:1-100.

and hydrochloric acids in a 1:2 molar ratio. Schiff reagents made

5. Hardonk MJ, van Duijn P. The mechanism of the Schiff reaction

as studied with histochemical model systems. J Histochem Cytochem
1964;12(10):748-51.

with thionyl chloride can be stored for a few years in securely capped
bottles (4, 9). Probably the most popular recipes for Schiffs reagent
(4, 7, 10) are those in which the sulfurous acid is formed by dissolving sodium metabisulfite in the dye solution and then adding
hydrochloric acid, allowing control of both the pH and sulfurous
acid content as well as avoiding the use of more hazardous
chemicals. Many vendors sell Schiffs reagent as a ready-made
solution. A survey of 20 material safety data sheets (MSDS) shows
that all were made with sodium metabisulfite, bisulfite or sulfite
(0.1-2.0%), and hydrochloric acid, with concentrations of basic
fuchsine ranging from 0.01 to 1.0%. The pH was reported in only

6. Robins JH, Abrams GD, Pinock JA. The structure of Schiff reagent
aldehyde adducts and the mechanism of the Schiff reaction as
determined by nuclear magnetic resonance spectroscopy. Can J Chem
1980;58:339-47.
7. Pearse AGE, Stoward PJ. Histochemistry, Theoretical and Applied, 4th
edn. Vol. 2. Analytical Technique. Edinburgh: Churchill-Livingstone,
1985 pp. 850-852.
8. Heinemann RL. Comparative spectroscopy of basic fuchsin, Schiff
reagent, and a formalin-Schiff derivative in relation to dye structure and
staining. Stain Technol 1970;45(4):165-72.
9. Kiernan JA. Histological and Histochemical Methods: Theory and
Practice. 4th ed. Bloxham, UK: Scion, 2008, pp. 224-227, 293-296.

seven of the MSDS, usually as a range such as 1.1-1.5 or 1.5-2.0.

10. Lillie RD, Fullmer HM. Histopathologic Technic and Practical

Histochemistry. 4th ed. New York: McGraw-Hill, 1976, pp. 305-308.

The two most frequent histochemical applications of Schiffs reagent

11. McManus JFA, Mowry RW. Staining Methods. Histologic and

Histochemical. New York: P. B. Hoeber, 1960 pp. 75, 126-128.

are the periodic acid-Schiff (PAS) method for neutral mucosubstances

and the Feulgen reaction, a specific stain for DNA. Investigations in
the 1950s indicated that maximum staining intensity was achieved at
pH 2.4 for the PAS method, whereas pH optima ranging from 2.3 to 4.3
Figure 2. Effects of pH on equilibria in solutions containing sulfite, bisulfite, sulfurous acid and sulfur dioxide (VII). Effects of H2SO3 and H+ on the protonation of Schiffs

were reported for the Feulgen technique (4). Current practice favors

reagent (VI, VIII, IX) and on the product of reaction with tissue-bound aldehyde groups (X, XI).

more acidic and therefore more stable solutions. The ambient pH

170 | special stains and h & E

References

12. Demalsy P, Callebaut M. Plain water as a rinsing agent preferable to

sulfurous acid after the Feulgen method. Stain Technol 1967;42:133-6.
13. Carson FL. Histotechnology. A Self-Instructional Text. 2nd ed. Chicago:
American Society of Clinical Pathologists, 1997, p. 114.
14. Churukian CJ. Manual of the Special Stains Laboratory (2000
edition). Rochester, NY: University of Rochester Medical Center, 2000,
pp. 52-55, 119-120; Also http://www.urmc.rochester.edu/path/zqu/
StainsManual/PERIODICACIDSCHIFFMETHOD.html (2009).

special stains and H & E

| 171

On Chemical Reactions and Staining Mechanisms

What is Giemsas stain and how does it color

blood cells, bacteria and chromosomes?

On Chemical Reactions and Staining Mechanisms

In most blood stains the anionic dye is eosin Y (tetrabromofluorescein),

which is easily purified. The stains of Leishman and of Macneal have
eosin B, which is often a mixture of related dyes.

Stains for Blood

Probably the first use of a staining solution containing methylene blue
(a cationic thiazine dye) and eosin (a red anionic xanthene dye) was
in 1888 by C. Chenzinsky, to stain malaria parasites in blood films.
F. Plehn in 1890 and E. Malachowski and D. L. Romanowsky in 1891
all independently described similar staining solutions that imparted a
variety of colors to erythrocytes, leukocytes and malaria parasites (1).
Notably, leukocyte nuclei were purple and the parasite nuclei were
bright red; these colors could not be obtained with any technique in
which the two dyes were used sequentially (2). Romanowskys thesis
(in Russian) and publication (in German) noted that the desired colors
were obtained only when the methylene blue was derived from a stock

Giemsas Formulation
Gustav Giemsa (1867-1948) endeavored to produce mixtures of dyes
that would reliably provide the Romanowsky color scheme (6). These
formulations were published in 1902-1904, with the most advanced
version being an aqueous solution containing a large excess of
azure II over eosin Y. Azure II was a deliberate mixture of unoxidized
methylene blue with azure I. Azure I was the name used at that time
for the dye now called azure B. Modern formulations of Giemsas
stain are made by mixing azure B eosinate with methylene blue and
are commercially available as Giemsa powder, which is dissolved in

solution old enough to have mold growing on its surface. The improved

a 50/50 mixture of methanol and glycerol to provide a concentrated

staining properties of ripened methylene blue solutions were studied

stock solution (9). For staining, the stock solution is diluted in water that

by Unna (3), who also published in 1891 and found that the aging

is buffered to a pH appropriate to the intended application pH 6.5-

process could conveniently be accelerated by heat and alkalinity.

7.0 for alcohol-fixed blood films, lower for sections of formaldehyde-

Artificially ripened or polychrome methylene blue was used by most

fixed tissues (2).

later inventors of blood stains, including L. Jenner (1899), B. Nocht

(1899), W. Leishman (1901), R. May & L. Grunwald (1902), and J. Wright
(1902) (1, 2). The products of polychroming, studied by Kehrmann
(4) and later by Homes and French (5), result from demethylation of
methylene blue and also from oxidation of the resulting dyes (Fig. 1).

Dyes Needed for the Romanowsky-Giemsa Effect

The compositions of blood stains were critically studied, using modern
chromatographic and analytical methods, in the 1970s, notably
by Paul Marshall in the UK (7, 8) and Dietrich Wittekind (2, 10) in

Staining solutions made by mixing solutions of polychrome methylene

Germany. These investigations revealed that the nuclear and all

blue and eosin are unstable because the oppositely charged dye ions

other colors constituting the Romanowsky-Giemsa effect could be

combine to form salts, known as azure eosinates, that are insoluble in

obtained with solutions containing only azure B and eosin Y. Pure

water. Azure eosinates are soluble in alcohols a mixture of methanol

azure B, made by direct synthesis rather than from polychromed

and glycerol is commonly used and they dissolve in water in the

methylene blue, has been commercially available (as chloride,

presence of an excess of either thiazine dye or eosin. Furthermore,

continued oxidation results in the deposition of other water-insoluble
products such as methylene violet (Bernthsen) (Fig. 1). For the
production of commercial blood stains the precipitated eosinate is
collected and dried (6, 7). Blood stains devised after 1902, such as

tetrafluoroborate or thiocyanate) since 1980 (11) and is used in a

Standardized Romanowsky-Giemsa method (12) in which a stable
methanolic solution of the eosinate is diluted with an aqueous buffer
to make the working staining solution.

Macneals tetrachrome (1922) and the formulation of Roe, Lillie &

Pure azure B is more expensive than the crude variety, and there

Wilcox (1941) made use of eosinates of azures A and B and methylene

is still considerable demand for Giemsas, Wrights, Macneals and

blue. Later work showed that all commercially produced thiazine dyes

related blood stains made from polychrome methylene blue and from

were mixtures of the compounds shown in Figure 1, with the nominal

impure azures A and B. The chemical compositions and staining

dye sometimes not the most abundant component (6, 8).

properties of these traditional mixtures have been shown to vary

172 | special stains and h & E

Figure 1. Dyes present in Giemsas and other blood stains. The arrows indicate the progressive demethylation and oxidation that constitute polychroming of
methylene blue.

special stains and H & E

| 173

On Chemical Reactions and Staining Mechanisms

On Chemical Reactions and Staining Mechanisms

Figure 2. Thin layer chromatography of

certified commercial samples of azure B
and Giemsas stain, on Unisil silica gel
plates. Solutions in methanol were applied
to the origin and development was in
butanol-water-acetic acid for 120 minutes.
Transcribed from cromatograms prepared
by Mr Matthew Frank in the Biological
Stain Commissions laboratory, Rochester
NY. Spots 3 and 4, both labeled azure A,
probably represent azure A and its isomer
sym-dimethylthionine respectively (7, 8).

among different suppliers and even among different lots from the

proteins are hemoglobin in erythrocytes and the major basic protein of

same supplier (7, 8). In the USA, samples of batches of methylene

eosinophil granules (17). The red coloration of malaria parasite nuclei

blue, azures A, B and C, methylene violet (Bernthsen) and powders

may be a due to a preponderance of basic proteins (histone) over

for Giemsas Jenners and Wrights stains are tested by the Biological

DNA in the nuclei of protozoa (18, 19) and some invertebrates (20).

Stain Commission (BSC) by assaying dye content, spectrophotometry,

thin-layer and high performance liquid chromatography and use in
standardized staining procedures (13). Satisfactory batches are
certified, and each bottle of dye powder carries a small additional
label, provided by the BSC, attesting to this fact. Certified dyes should
always be used for preparing solutions of Romanowsky-Giemsa
stains. Azure B, the single most important ingredient of these stains,
does not have to be the ultra-pure directly synthesized dye to meet
the BSCs requirements for certification (Fig. 2).

Giemsa for Tissue Sections

The expected colors in cells of blood or bone marrow are seen only
in alcohol-fixed films or smears and when the diluted staining mixture
is close to neutrality - usually pH 6.8. Wrights or Leishmans stain
is usually used, being allowed to act for about 3 minutes. Giemsas
stain, which is much more stable when diluted, can be allowed to
act for 15-45 minutes. After fixation with formaldehyde, which reacts
with amino and other basic groups, nearly everything stains blue
at pH 6.8. Paraffin sections can be stained with Giemsa at pH 4 to

Staining Mechanisms

5. The sections show blue (not purple) nuclei, pink erythrocytes,

Studies of the interaction of DNA with azure B (14, 15) indicate that

cytoplasm and collagen, and metachromatic (red-purple) mast cells

in solution the dye cations are present as dimers, each with the two

and cartilage. The colors are not the same as those seen in alcohol-

planar ring systems held together by van der Waals forces. The dimers

fixed blood films (21-23). Giemsa is often used to stain bacteria in

are attracted to the negatively charged phosphate groups of DNA and

sections of formaldehyde-fixed gastric biopsies. The organisms are

adhere to the macromolecule as a result of hydrophobic interactions

dark blue against an unstained or light pink background (23). The

with the purine and pyrimidine rings of the DNA bases. Attraction

substrates of staining are presumably bacterial DNA and RNA, which

of eosin anions by un-neutralized positive charges of bound azure

differ from their animal counterparts in not being associated with basic

B dimers changes the color of the stained DNA from blue to purple

proteins (24). The teichoic acids of the cell walls of Gram-positive

(16). Eosin is also attracted to proteins with excess protonated amino

bacteria probably also bind cationic dyes, and the phosphate groups

and guanidino groups (mainly side chains of lysine and arginine) over

of lipopolysaccharides can be expected to have a similar role in Gram-

ionized carboxy groups (mainly glutamic and aspartic acids); these

negative organisms.

174 | special stains and h & E

Figure 3. Karyogram of GTG-banded human chromosomes ( ) (46 chromosomes, including XY): GTG = G-banding with trypsin-Giemsa. Courtesy of Dr. Stefan Mueller,
Ludwig-Maximilians-University, Munich, Germany.

special stains and H & E

| 175

Chapter 20 | H&E Staining

On Chemical Reactions and Staining Mechanisms

Gary W. Gill, CT(ASCP)

Chromosome Banding
Giemsa (G-banding) is the stain most commonly used to display
transverse bands in mitotic (metaphase) chromosomes, having
largely replaced quinacrine and other fluorochromes (Q-banding)
that were formerly used for this purpose. There are various empirically
developed G-banding techniques in which chromosome preparations
typically cell cultures that have been arrested in metaphase, spread
onto slides, and fixed in 3:1 methanol-acetic acid are subjected to
pre-treatments before staining with Giemsa (25, 26). The dye cations
in the Giemsa mixture (azure B and methylene blue) are attracted by
the phosphate anions of DNA and are then held more closely by van
der Waals forces to regions rich in adenine and thymine. Regions of
a chromosome rich in guanine-cytosine pairs do not bind the blue
dyes. The pretreatment determines the type of banding pattern. In
typical G-banding the preparations are treated either with a sodium
chloride and citrate solution at 60C for one hour or with trypsin for 10
seconds at room temperature before staining in dilute Giemsa at pH
6.8 for 45 minutes (Fig. 3). R-banding is a method that shows bands
that are complementary to the G-bands; that is, the dark and light
regions of the G-banded chromosome are reversed. The pretreatment
for R-banding is with 1.0 M NaH2PO4 (pH about 4) at 88C for 20
minutes, and the staining is with a more concentrated Giemsa
solution, for 10 minutes (27). G-banding is the method usually used for
identifying chromosomes. R-banding is useful for showing the ends of
chromosomes (telomeres). A variant known as T-banding emphasizes
these terminal segments; after treatment with a phosphate buffer,
pH 5.1 at 87C for 20-60 minutes the preparations are stained in
Giemsa, also at 87C (25). The histochemical rationales of the various
pretreatments have not been systematically investigated (26).

7. Marshall PN. Romanowsky-type stains in haematology. Histochem J

1978;10:1-29.
8. Marshall PN. Romanowsky staining: state of the art and ideal
techniques. In: Koopke JA, ed. Differential Leukocyte Counting. Skokie,
IL: American College of Pathologists, 1979:205-216.

What is the difference between progressive

vs. regressive hematoxylin staining?

Are there reasons to prefer water or alcohol

as the solvent for eosin formulations?

9. Sanderson JB. Biological Microtechnique. Oxford: BIOS Scientific

Publications & Royal Microscopical Society, 1994: 176.

Progressive hematoxylin stains color primarily chromatin and to a

I prefer alcohol-based eosin formulations: 1) they are chemically

10. Green FJ. The Sigma-Aldrich Handbook of Stains, Dyes and Indicators.
Milwaukee, Wisconsin: Aldrich Chemical Company, 1990: 109-122,
388-389, 413-414, 448-450, 734-736.

much less extent cytoplasm to the desired optical density, regardless

more stable 2) they minimize, if not eliminate entirely, the unpre-

of the length of staining time. Regressive hematoxylin stains over-

dictable effects of various impurities such as water-soluble salts

11. Wittekind DH. Romanowsky-Giemsa stains. In: Horobin RW, Kiernan

JA, eds. Conns Biological Stains. 10th ed. Oxford: BIOS, 2002: 303-312.

stain chromatin and cytoplasm and require subsequent immersion

that in water may interfere with dye uptake, and 3) they tend to stain

in dilute acid to pull out the excess color from the chromatin and

more slowly than water-based formulations (promotes a wider range

12. Wittekind DH, Kretschmer V. On the nature of RomanowskyGiemsa staining and the Romanowsky-Giemsa effect. II. A revised
Romanowsky-Giemsa staining procedure. Histochem J 1987;19:399-401.

cytoplasm (Table 1). If differentiation is omitted or incomplete, residual

of shades of eosin colors).

13. Penney DP, Powers JM, Frank M, Churukian C. Analysis and testing
of biological stains the Biological Stain Commission procedures.
Biotech Histochem 2002;77(5-6):237-275.

the uptake of eosin entirely.

14. Muller-Walz R, Zimmermann HW. Uber Romanowsky-Farbstoffe und

den Romanowsky-Giemsa-Effekt. 4. Mitteilung: Bindung von Azur B an
DNA. Histochemistry 1987;87:157-172.
15. Friedrich K, Seiffert W, Zimmermann HW. Romanowsky dyes and
Romanowsky-Giemsa effect. 5. Structural investigations of the purple
DNA-AB-EY dye complexes of Romanowsky-Giemsa staining.
Histochemistry 1990;93:247-256.
16. Horobin RW, Walter KJ. Understanding Romanowsky staining.
I. The Romanowsky-Giemsa effect in blood smears. Histochemistry
1987;86:331-336.
17. Weller PF, Ackerman SJ, Smith JA. Eosinophil granule cationic
proteins: major basic protein is distinct from the smaller subunit of
eosinophil peroxidase. J Leukocyte Biol 1988;43(1):1-4.
18. Khanna DR, Yadav PR. Biology of Protozoa. New Delhi: Discovery
Publishing House, 2004:21-85.
19. Gill J, Yogavel M, Kumar A, et al. Crystal structure of malaria
parasite nucleosome assembly protein. Distinct modes of protein
localization and histone recognition. J Biol Chem 2009;284(15):
10076-10087.
20. Pasternak J, Barrell R. Cytophotometric study of nuclear proteins
during embryogenesis in two nematode species, Ascaric lumbricoides
and Panagrellus silusiae. J Exp Zool 1976;198:217-224.
21. Lillie RD, Fullmer HM. Histopathologic Technic and Practical Histochemistry.
4th ed. New York: McGraw-Hill, 1976: 193-197.

References
1. Woronzoff-Dashkoff KPK. The Ehrlich-Chenzinzky-Plehn-MalachowskiRomanowsky-Nocht-Jenner-May-Grunwald-Leishman-Reuter-WrightGiemsa-Lillie-Roe-Wilcox stain. The mystery unfolds. Clin Lab Med
1993;13(4):759-771.
2. Wittekind DH. On the nature of Romanowsky-Giemsa staining and its
significance for cytochemistry and histochemistry: an overall review.
Histochem J 1983;15:1029-1047.

22. Kiernan JA. Histological and Histochemical Methods: Theory and

Practice. 4th ed. Bloxham, UK: Scion, 2008: 159.
23. Carson FL. Histotechnology. A Self-Instructional Text. 2nd ed. Chicago:
American Society of Clinical Pathologists, 1997: 106-107, 190-191.
24. Robinow C, Kellenberger E. The bacterial nucleoid revisited. Microbiol
Rev 1994;58(2):211-232.

3. Unna PC. Uber die Reifung unsrer Farbstoffe. Z wiss Mikr 1891;8:475-487.

25. Gosden JR, ed. Chromosome Analysis Protocols. Totoya, NJ: Humana
Press, 1994: 59-81.

4. Kehrmann F. Ueber Methylen-azur. Ber Deutsch Chem Ges 1906;39(2):

1403-1408.

26. Sumner AT. Chromosomes. Organization and Function. Oxford: Blackwell,

2003.

5. Holmes WC, French RW. The oxidation products of methylene blue.

Stain Technol 1926;1:17-26.

27. Sehested J. A simple method for R banding of human chromosomes,

showing pH-dependent connection between R and G bands.
Humangenetik 1974;21:55-58.

6. Barcia JJ. The Giemsa stain: its history and applications. Int J Surg Path
2007;15(3):292-296.

176 | special stains and h & E

hematoxylin visually obscures fine chromatin detail and can prevent

respectively, and 25% ethylene glycol. They are progressive stains

Is there a simple way to perform quality

assurance (QA) on hematoxylin and eosin
stains before using a batch for the first time?

that can be applied for many minutes without overstaining and without

Yes. Whether buying or making hematoxylin eosin solutions, one

differentiation in a dilute acid bath. Harris hematoxylin contains 5 gm

cannot be absolutely certain the product will perform. Apart from

hematoxylin per liter of water. It overstains within minutes and requires

unsound methods, limitations in ingredients, incorrect formulations

differential extraction in dilute HCl to decolorize the cytoplasm

(e.g., precipitated mordant crystals in commercial Harris hematoxylin

(differentiation) and to remove excess hematoxylin from chromatin.

formulations), and errors in formulating (e.g., weighing out too much

Figure 1 illustrates the difference between the two approaches.

oxidizing agent) can contribute to unsatisfactory results. It doesnt

Gill hematoxylins No. 1 and 2 contain 2 and 4 gm hematoxylin per liter,

happen often, but it does happen. Regulatory documentation does

Do you have a preference for progressive

or regressive hematoxylin staining?

not guarantee efficacy.

I prefer progressive hematoxylin staining because it does not require

invaluable probes to: 1) determine the performance of each new lot

differentiation. Under- or over-differentiation can produce over-

of stain, 2) select suitable staining times, 3) find out how many slides

staining or understaining. Depending on the degree of timing control

can be stained satisfactorily by a given volume of each stain, 4) learn

exercised in a given laboratory, the results may be satisfactory one

when rinses should be changed, and 5) troubleshoot whether a given

day, hyperchromatic another day, and hypochromatic the next.

stain already in use is the cause of an observed staining problem.

Extreme hyperchromasia can block entirely the uptake of eosin so

Once experience imparts confidence to selected staining times, stain

that H&E becomes simply H.

and rinse change schedules, the use of control sections or smears is

Formalin-fixed tissue sections or alcohol-fixed buccal smears are

not necessary for the remainder of the life of the particular stain that

What is the difference between differentiation

and bluing?
Differentiation and bluing (blueing, if you prefer the English spelling)
are essential to satisfactory staining by hematoxylins. Differentiation

has been validated. However, control preparations should be used

when new containers of the same stain with different lot numbers are
opened to confirm that the stain does indeed perform as expected.
Manufacturers occasionally make bad batches of stain.

is used only with regressive hematoxylin formulations, while bluing is

Simply looking at one of the first slides stained daily and initialing a

used with both regressive and progressive hematoxylin formulations.

stain quality log sheet is of no value if a laboratory has not defined its

Differentiation effects quantitative changes; bluing, qualitative.

standards. It is not uncommon to see such sheets dutifully maintained

See Table 2.

and also to see unsatisfactory staining results.

special stains and H & E

| 177